@phdthesis{Segebarth2021,
  author    = {Segebarth, Dennis},
  title     = {Evaluation and validation of deep learning strategies for bioimage analyses},
  doi       = {10.25972/OPUS-24372},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-243728},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2021},
  abstract  = {Significant advances in fluorescence imaging techniques enable life scientists today to gain insights into biological systems at an unprecedented scale. The interpretation of image features in such bioimage datasets and their subsequent quantitative analysis is referred to as bioimage analysis. A substantial proportion of bioimage analyses is still performed manually by a human expert - a tedious process that is long known to be subjective. Particularly in tasks that require the annotation of image features with a low signal-to-noise ratio, like in fluorescence images of tissue samples, the inter-rater agreement drops. However, like any other scientific analysis, also bioimage analysis has to meet the general quality criteria of quantitative research, which are objectivity, reliability, and validity. Thus, the automation of bioimage analysis with computer-aided approaches is highly desirable. Albeit conventional hard-coded algorithms are fully unbiased, a human user has to set its respective feature extraction parameters. Thus, also these approaches can be considered subjective. Recently, deep learning (DL) has enabled impressive advances in computer vision research. The predominant difference between DL and conventional algorithms is the capability of DL models to learn the respective task on base of an annotated training dataset, instead of following user-defined rules for feature extraction. This thesis hypothesized that DL can be used to increase the objectivity, reliability, and validity of bioimage analyses, thus going beyond mere automation. However, in absence of ground truth annotations, DL models have to be trained on manual and thus subjective annotations, which could cause the model to incorporate such a bias. Moreover, model training is stochastic and even training on the same data could result in models with divergent outputs. Consequently, both the training on subjective annotations and the model-to-model variability could impair the quality of DL-based bioimage analyses. This thesis systematically assessed the impacts of these two limitations experimentally by analyzing fluorescence signals of a protein called cFOS in mouse brain sections. Since the abundance of cFOS correlates with mouse behavior, behavioral analyses could be used for cross-validation of the bioimage analysis results. Furthermore, this thesis showed that pooling the input of multiple human experts during model training and integration of multiple trained models in a model ensemble can mitigate the impact of these limitations. In summary, the present study establishes guidelines for how DL can be used to increase the general quality of bioimage analyses.},
  subject      = {Deeplearning},
  language  = {en}
}
@phdthesis{Hu2021,
  author    = {Hu, Xiawei},
  title     = {Role of claudin-12 in neuronal barriers in painful murine and human neuropathy},
  doi       = {10.25972/OPUS-20806},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-208065},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2021},
  abstract  = {In peripheral nervous system (PNS), the blood-nerve barrier (BNB) and myelin barrier (MB) are important physiological fences for maintaining the environment for axons, Schwann cells and other associated cells within peripheral nerves. The perineurium surrounding the nerves and endoneurial vessels nourishing the nerves compose the BNB. Schwann cells wrapping around neurons form the MB. Destruction or malfunction of the barriers has been postulated as an initial step in the development of pathologic conditions concerning human peripheral nerves, such as traumatic neuropathy and the disease of chronic inflammatory demyelination polyneuropathy (CIDP). Tight junction proteins (TJPs) are intercellular junctions building the microstructure of barriers. They play a key role in tightly connecting adjacent cells, controlling the passage of ions, water and other molecules via the paracellular pathway, and maintaining the cell polarity. Among the family of TJPs, claudins are the major structural components which form the backbone of TJs. Certain key TJPs [e.g. claudins (claudin-1, -5, -19, occludin, zona occludens (ZO-1)] have been identified in neural barriers and explored for therapeutic targets. The expression of Cldn12 gene has been documented in human/rodent tibial nerves, spinal cord and DRG. However, the role of claudin-12 in PNS is unknown. In the present study, we firstly found a loss of claudin-12 immunoreactivity (IR) in male or postmenopausal female patients with painful CIDP or non-inflammatory polyneuropathy (PNP). Then, we utilized male and female Cldn12-KO mice and the chronic constriction injury (CCI) model. Cldn12 mRNA and IR were reduced in WT mice after nerve injury. Deletion of Cldn12 via general knockout (KO) induced mechanical allodynia at baseline level and after CCI in time-dependent manner in male mice. KO of Cldn12 in males resulted in loss of small axons, perineurial barrier and MB breakdown, as well as TJP complex disruption with claudin-1, -19 and Pmp22 reduction. Moreover, local Cldn12 siRNA application mimicked mechanical allodynia and MB breakdown. In conclusion, claudin-12 deficiency is associated with painful CIDP/non-inflammatory PNP. Claudin-12 is a regulatory TJP crucial for mechanical nociception, perineurial barrier and MB integrity, and proper TJP composition in mice. Therefore, further investigating the functions of claudin-12 and its mechanism is important to prompt the development of new therapeutic approaches for painful neuropathies.},
  language  = {en}
}
@phdthesis{Mohammadi2019,
  author    = {Mohammadi, Milad},
  title     = {Role of oxidized phospholipids in inflammatory pain},
  doi       = {10.25972/OPUS-19240},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-192402},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2019},
  abstract  = {Introduction: During inflammation, reactive oxygen species (ROS) such as Hydrogen peroxide accumulate at the inflammation site and by oxidizing lipids, they produce metabolites such as 4-hydroxynonenal (4-HNE) and oxidized phospholipids (OxPLs). Transient receptor potential ankyrin 1 (TRPA1) and vanilloid 1 (TRPV1) are ligand gated ion channels that are expressed on nociceptors and their activation elicits pain. Hydrogen peroxide and 4-HNE are endogenous ligands for TRPA1 and their role in inflammatory pain conditions has been shown. OxPLs play a major pro-inflammatory role in many pathologies including atherosclerosis and multiple sclerosis. E06/T15 is a mouse IgM mAb that specifically binds oxidized phosphatidylcholine. D-4F is an apolipoprotein A-I mimetic peptide with a very high affinity for OxPLs and possess anti-inflammatory properties. E06 mAb and D-4F peptide protect against OxPLs-induced damage in atherosclerosis in vivo. Methods: To investigate the role of ROS and their metabolites in inflammatory pain, I utilized a combination of diverse and complex behavioral pain measurements and binding assays. I examined E06 mAb and D-4F as local treatment options for hypersensitivity evoked by endogenous and exogenous activators of TRPA1 and TRPV1 as well as in inflammatory and OxPL-induced pain models in vivo. 4-HNE, hydrogen peroxide as ROS source and mustard oil (AITC) were used to activate TRPA1, while capsaicin was used to activate TRPV1. Results: Intraplantar injection of oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine (OxPAPC) into rats' hind paw elicited thermal and mechanical hypersensitivity. Genetic and pharmacological evidence in vivo confirmed the role of TRPA1 in OxPLs-induced hypersensitivity. OxPLs formation increased in complete Freund's adjuvant (CFA)-induced inflamed rats' paw. E06 mAb and D-4F prevented OxPAPC-induced mechanical and thermal hypersensitivity (hyperalgesia) as well as CFA-induced mechanical hypersensitivity. Also, all irritants induced thermal and mechanical hypersensitivity as well as affective-emotional responses and spontaneous nocifensive behaviors. E06 mAb blocked prolonged mechanical hypersensitivity by all but hydrogen peroxide. In parallel, D-4F prevented mechanical hypersensitivity induced by all irritants as well as thermal hypersensitivity induced by capsaicin and 4-HNE. In addition, competitive binding assays showed that all TRPA1/V1 agonists induced prolonged formation of OxPLs in the paw tissue explaining the anti-nociceptive properties of E06 mAb and D-4F. Finally, the potential of gait analysis as a readout for non-provoked pain behavioral measurements were examined. Conclusion and implications: OxPLs were characterized as novel targets in inflammatory pain. Treatment with the monoclonal antibody E06 or apolipoprotein A-I mimetic peptide D-4F are suggested as potential inflammatory pain medications. OxPLs' role in neuropathic pain is yet to be investigated.},
  language  = {en}
}
@phdthesis{Schulte2023,
  author    = {Schulte, Annemarie},
  title     = {\(In\) \(vitro\) reprogramming of glial cells from adult dorsal root ganglia into nociceptor-like neurons},
  doi       = {10.25972/OPUS-30311},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-303110},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Plexus injury often occurs after motor vehicle accidents and results in lifelong disability with severe neuropathic pain. Surgical treatment can partially restore motor functions, but sensory loss and neuropathic pain persist. Regenerative medicine concepts, such as cell replacement therapies for restoring dorsal root ganglia (DRG) function, set high expectations. However, up to now, it is unclear which DRG cell types are affected by nerve injury and can be targeted in regenerative medicine approaches. This study followed the hypothesis that satellite glial cells (SGCs) might be a suitable endogenous cell source for regenerative medicine concepts in the DRG. SGCs originate from the same neural crest-derived cell lineage as sensory neurons, making them attractive for neural repair strategies in the peripheral nervous system. Our hypothesis was investigated on three levels of experimentation. First, we asked whether adult SGCs have the potential of sensory neuron precursors and can be reprogrammed into sensory neurons in vitro. We found that adult mouse DRG harbor SGC-like cells that can still dedifferentiate into progenitor-like cells. Surprisingly, expression of the early developmental transcription factors Neurog1 and Neurog2 was sufficient to induce neuronal and glial cell phenotypes. In the presence of nerve growth factor, induced neurons developed a nociceptor-like phenotype expressing functional nociceptor markers, such as the ion channels TrpA1, TrpV1 and NaV1.9. In a second set of experiments, we used a rat model for peripheral nerve injury to look for changes in the DRG cell composition. Using an unbiased deep learning-based approach for cell analysis, we found that cellular plasticity responses after nerve injury activate SGCs in the whole DRG. However, neither injury-induced neuronal death nor gliosis was observed. Finally, we asked whether a severe nerve injury changed the cell composition in the human DRG. For this, a cohort of 13 patients with brachial plexus injury was investigated. Surprisingly, in about half of all patients, the injury-affected DRG showed no characteristic DRG tissue. The complete entity of neurons, satellite cells, and axons was lost and fully replaced by mesodermal/connective tissue. In the other half of the patients, the basic cellular entity of the DRG was well preserved. Objective deep learning-based analysis of large-scale bioimages of the "intact" DRG showed no loss of neurons and no signs of gliosis. This study suggests that concepts for regenerative medicine for restoring DRG function need at least two translational research directions: reafferentation of existing DRG units or full replacement of the entire multicellular DRG structure. For DRG replacement, SGCs of the adult DRG are an attractive endogenous cell source, as the multicellular DRG units could possibly be rebuilt by transdifferentiating neural crest-derived sensory progenitor cells into peripheral sensory neurons and glial cells using Neurog1 and Neurog2.},
  subject      = {Spinalganglion},
  language  = {en}
}
@phdthesis{Brych2022,
  author    = {Brych, Mareike Kimberly},
  title     = {How movements and cognition interact: An investigation of spontaneous blinks},
  doi       = {10.25972/OPUS-26737},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-267376},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2022},
  abstract  = {During natural behavior, cognitive processes constantly coincide with body movements such as head or eye movements or blinks. However, during experimental investigations of cognitive processes, movements are often highly restricted which is rather unnatural. In order to improve our understanding of natural behavior, this thesis investigates the interaction between cognition and movements by focusing on spontaneous blinks, which naturally interact with other body movements. Spontaneous blinks are inevitably connected to vision as they shut out incoming visual information. Both sensory-based and cognitive factors, for example, stimulus occurrence and evaluation, were reported to influence blink behavior. Our first study investigated if such influences are comparable for visual and non-visual input. The chosen experimental design allowed dissociating sensory-driven and cognitive influences, which then could be compared between the visual and auditory domain. Our results show that blinks are more strongly modulated during passive observation of visual input compared to auditory input. This modulation is however enhanced for both input modalities by an increased attentional demand. In addition, the cognitively defined meaning of a stimulus changes blink latency independent of the sensory domain. Overall, our findings show that spontaneous blinks and cognitive processes are linked beyond vision. Moreover, the underlying cognitive processes that influence blinks are largely the same across different sensory input indicating that blinks are profoundly integrated into our system. When investigating natural behavior, it is important to consider that movements rarely occur in isolation, but are executed side by side. As these movements interact and have a link to cognitive processes, the complexity of our system increases. In order to take this complexity into account, the second part of the experimental research focused on movement interactions, more specifically on the interactions between blinks, pupil size and speaking. Our results reveal that speech-related motor activity increases blink rate and pupil size as well as modulates blink timing. This is in line with previous research that described a relation between different body and eye movements. Importantly, each bodily-induced change in eye movements affects visual information intake. Therefore, different movements can be tightly linked to perceptual processes through complex interactions. Altogether, the work of this thesis provides rich evidence that movements and cognitive processes are deeply intertwined. Therefore, movements should be seen as an integral part of our system. Taking the relevance of movements and their interactions into account during experimental investigations is necessary in order to reveal a more realistic and complete picture of human natural behavior.},
  subject      = {Kognition},
  language  = {en}
}
@phdthesis{Klinke2022,
  author    = {Klinke, Christopher Matthias},
  title     = {Experimental investigation of the effect of distal stress induction on threat conditioning in humans},
  doi       = {10.25972/OPUS-22556},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-225562},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2022},
  abstract  = {Stress constitutes a major risk factor for the development of psychiatric disorders, such as PTSD and anxiety disorders, by shifting the brain into a state of sensitization and makes it more vulnerable when being exposed to further aversive events. This was experimentally in-vestigated in rodents by examining the effect of a distal stress induction on threat conditioning, where stress impaired extinction learning and caused spontaneous recovery. However, this effect has never been experimentally investigated in humans, so far. Thus, the aim of this dissertation was to investigate the effect of distal stress on threat conditioning in humans. Therefore, two subsequent studies were conducted. For both studies, the threat conditioning paradigm comprised threat acquisition, extinction learning, and re-extinction. In the threat acquisition phase, two geometrical shapes were used as conditioned stimulus (CS), from which one (CS+) was paired with a painful electric stimulus (unconditioned stimulus, US), but not the other one (CS-). During extinction learning 24 h later and re-extinction seventeen days later, CSs were again presented but without any US delivery. In Study 1, 69 participants underwent either a stress (socially evaluated cold pressor test; SECPT) or sham protocol 10 days prior to threat conditioning. Furthermore, context effects were examined by placing the stress protocol in the same context (context-A stress, and sham group) or a different context (context-B stress group) than conditioning. Results revealed that the context-A, but not context-B, stress group displayed impaired safety learning (i.e. potenti-ation towards CS-) for startle response during threat acquisition. Moreover, the same stress group showed impaired threat extinction, evident in sustained CS discrimination in valence and arousal ratings during extinction learning, and memory recall. In sum, distal stress on the one hand impaired safety learning during threat conditioning on a level of startle response. On the other hand, stress impaired threat extinction on a level of ratings. Noteworthy, the effect of distal stress was only found when the stressor was placed in the same context as later threat learning. Hence, suggesting that the combination of stressor and stressor-associated context exerted the effect on threat extinction. In Study 2, it was examined if distal stress induction could also have an impact on threat and extinction processes without the necessity of context association. Therefore, the same stress (n = 45) or sham protocol (n = 44) as in Study 1 was conducted in a different context than and 24 h prior to a threat conditioning paradigm. Similar to Study 1, weakened extinction learning was found in fear ratings for the stress (vs. sham) group, which was indicated by persistent CS+/CS- differentiation after the first block of extinction trials. Alterations in safety learning towards the CS- during threat acquisition were only supported by significant correlations between stress measures on the stress day and conditioned startle response of the CS- during acquisition. Taken together, in two subsequent studies this dissertation provided first evidence of impaired threat extinction after distal stress induction in humans. Furthermore, impairments in safety learning, as can be observed in PTSD, were additionally demonstrated. Interestingly, the effects were boosted and more profound when associating the stressor to the later learning context. These results have clinical implications as they can be translated to the notion that prior stress exposure makes an individual more vulnerable for later aversive events.},
  subject      = {Stress},
  language  = {en}
}
@phdthesis{Mortimer2021,
  author    = {Mortimer, Niall Patrick},
  title     = {ADHD Genetics in Mouse and Man},
  doi       = {10.25972/OPUS-23626},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-236265},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2021},
  abstract  = {Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder with an estimated heritability of around 70\%. In order to fully understand ADHD biology it is necessary to incorporate multiple different types of research. In this thesis, both human and animal model research is described as both lines of research are required to elucidate the aetiology of ADHD and development new treatments. The role of a single gene, Adhesion G protein-coupled receptor L3 (ADGRL3) was investigated using a knockout mouse model. ADGRL3 has putative roles in neuronal migration and synapse function. Various polymorphisms in ADGRL3 have been linked with an increased risk of attention deficit/hyperactivity disorder (ADHD) in human studies. Adgrl3-deficient mice were examined across multiple behavioural domains related to ADHD: locomotive activity, visuospatial and recognition memory, gait impulsivity, aggression, sociability and anxiety-like behaviour. The transcriptomic alterations caused by Adgrl3-depletion were analysed by RNA-sequencing of three ADHD-relevant brain regions: prefrontal cortex (PFC), hippocampus and striatum. Increased locomotive activity in Adgrl3-/- mice was observed across all tests with the specific gait analysis revealing subtle gait abnormalities. Spatial memory and learning domains were also impaired in these mice. Increased levels of impulsivity and sociability accompanying decreased aggression were also detected. None of these alterations were observed in Adgrl3+/- mice. The numbers of genes found to exhibit differential expression was relatively small in all brain regions sequenced. The absence of large scale gene expression dysregulation indicates a specific pathway of action, rather than a broad neurobiological perturbation. The PFC had the greatest number of differentially expressed genes and gene-set analysis of differential expression in this brain region detected a number of ADHD-relevant pathways including dopaminergic synapses as well as cocaine and amphetamine addiction. The most dysregulated gene in the PFC was Slc6a3 which codes for the dopamine transporter, a molecule vital to current pharmacological treatment of ADHD. The behavioural and transcriptomic results described in this thesis further validate Adgrl3 constitutive knockout mice as an experimental model of ADHD and provide neuroanatomical targets for future studies involving ADGRL3 modified animal models. The study of ADHD risk genes such as ADGRL3 requires the gene to be first identified using human studies. These studies may be genome based such as genome wide association studies (GWAS) or transcriptome based using microarray or RNA sequencing technology. To explore ADHD biology in humans the research described in this thesis includes both GWAS and trancriptomic data. A two-step transcriptome profiling was performed in peripheral blood mononuclear cells (PBMCs) of 143 ADHD subjects and 169 healthy controls. We combined GWAS and expression data in an expression-based Polygenic Risk Score (PRS) analysis in a total sample of 879 ADHD cases and 1919 controls from three different datasets. Through this exploratory study we found eight differentially expressed genes in ADHD and no support for the genetic background of the disorder playing a role in the aberrant expression levels identified. These results highlight promising candidate genes and gene pathways for ADHD and support the use of peripheral tissues to assess gene expression signatures for ADHD. This thesis illustrates how both human and animal model research is required to increase our understanding of ADHD. The animal models provide biological insight into the targets identified in human studies and may themselves provide further relevant gene targets. Only by combining research from disparate sources can we develop the thorough understanding on ADHD biology required for treatment development, which is the ultimate goal of translational science research.},
  language  = {en}
}
@phdthesis{HuttererneeHerzog2024,
  author    = {Hutterer, n{\´e}e Herzog, Katharina},
  title     = {Treatment-like use of discrimination training to reduce generalization of conditioned fear},
  doi       = {10.25972/OPUS-31728},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-317286},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {Anxiety patients overgeneralize fear, also because of an inability to perceptually discriminate threat and safety signals. Therefore, some studies have developed discrimination training that successfully reduced the occurrence of fear generalization. The present work is the first to take a treatment-like approach by using discrimination training after generalization has occurred. Therefore, two studies were conducted with healthy participants using the same fear conditioning and generalization paradigm, with two faces as conditioned stimuli (CSs), and four facial morphs between CSs as generalization stimuli (GSs). Only one face (CS+) was followed by a loud scream (unconditioned stimulus, US). In Study 1, participants underwent either fear-relevant (discriminating faces) or fear-irrelevant discrimination training (discriminating width of lines) or a non-discriminative control training between the two generalization tests, each with or without feedback (n = 20 each). Generalization of US expectancy was reduced more effectively by fear-relevant compared to fear-irrelevant discrimination training. However, neither discrimination training was more effective than non-discriminative control training. Moreover, feedback reduced generalization of US expectancy only in discrimination training. Study 2 was designed to replicate the effects of the discrimination-training conditions in a large sample (N = 244) and examine their benefits in individuals at risk for anxiety disorders. Again, feedback reduced fear generalization particularly well for US expectancy. Fear relevance was not confirmed to be particularly fear-reducing in healthy participants, but may enhance training effects in individuals at risk of anxiety disorder. In summary, this work provides evidence that existing fear generalization can be reduced by discrimination training, likely involving several (higher-level) processes besides perceptual discrimination (e.g., motivational mechanisms in feedback conditions). Its use may be promising as part of individualized therapy for patients with difficulty discriminating similar stimuli.},
  subject      = {Furcht},
  language  = {en}
}